Generalizing Movement Primitives to New Situations

Jens Lundell, Murtaza Hazara*, Ville Kyrki

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

4 Citations (Scopus)
177 Downloads (Pure)

Abstract

Although motor primitives (MPs) have been studied extensively, much less attention has been devoted to studying their generalization to new situations. To cope with varying conditions, a MP’s policy encoding must support generalization over task parameters to avoid learning separate primitives for each condition. Local and linear parameterized models have been proposed to interpolate over task parameters to provide limited generalization. In this paper, we present a global parametric motion primitive (GPDMP) which allows generalization beyond local or linear models. Primitives are modelled using a linear basis function model with global non-linear basis functions. The model is constructed from initial non-parametric primitives found using a single human demonstration and subsequent episodes of reinforcement learning to adapt the demonstrated skill to other task parameters. The initial models are then used to optimize the parameters of the global parametric model. Experiments with a ball-in-a-cup task with varying string lengths show that GPDMP allows greatly improved extrapolation compared to earlier local or linear models.
Original languageEnglish
Title of host publicationTowards Autonomous Robotic Systems - 18th Annual Conference, TAROS 2017, Proceedings
Pages16-31
Number of pages16
Volume10454 LNAI
ISBN (Electronic)978-3-319-64107-2
DOIs
Publication statusPublished - 2017
MoE publication typeA4 Article in a conference publication
EventTowards Autonomous Robotic Systems Conference - Guildford, United Kingdom
Duration: 19 Jul 201721 Jul 2017
Conference number: 18

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10454 LNAI
ISSN (Print)03029743
ISSN (Electronic)16113349

Conference

ConferenceTowards Autonomous Robotic Systems Conference
Abbreviated titleTAROS
CountryUnited Kingdom
CityGuildford
Period19/07/201721/07/2017

Keywords

  • learning from demonstration
  • generalization
  • global parametric model
  • ball-in-a-cup

Fingerprint

Dive into the research topics of 'Generalizing Movement Primitives to New Situations'. Together they form a unique fingerprint.

Cite this